skillpack-creator

# Skillpack Creator ## Overview Turn a successful task into a reusable SkillPack. Extract the stable workflow, decide what belongs in a local skill versus pack-level prompts, generate the pack structure, and package it only after the workflow is explicit and repeatable. ## Workflow ### 1. Normalize the source task Start by reducing the finished task into a clean execution spec: - Capture the user goal and the concrete deliverable. - Capture the final successful workflow, not the full exploratory transcript. - List required skills, tools, files, secrets, and environment assumptions. - Separate deterministic steps from heuristic steps. - Remove dead ends, retries, and one-off debugging noise. If the prior work is still ambiguous, ask for the missing stable facts or infer only the pieces that are low risk. ### 2. Decide what the pack should contain Use this split: - Put reusable procedural knowledge in a local skill under the target pack's `skills/`. - Put repeated shell or file-generation logic into scripts inside that local skill when reliability matters. - Put detailed schemas, API notes, or conventions into reference files. - Put 1 to 3 pack-level prompts in `skillpack.json` as the pack's user-facing entry points. Do not treat `prompts` as a strict workflow engine. In this codebase they are starter inputs for the UI, not a DAG or state machine. Read `references/skillpack-format.md` when you need the exact pack semantics. ### 3. Create the pack specification Befo...